Influence of maternal effects on heritability of wing length in Drosophila melanogaster. 

Stanić, Snežana¹, and Dragoslav Marinković².  ¹Faculty of Science, 34000 Kragujevac, and ²Faculty of Biology, 11000 Belgrade, Yugoslavia. 

      Heritability in wider sense is a useful parameter if we are interested in the genotype and environmental influences in formation of  phenotypic differences between individuals of a  population. In the paper of Stanić and Marinković (1999), it has been shown that heritability of wing length in wider sense is 0.26 in the female sex and 0.83 in the male sex of Drosophila melanogaster individuals. If we try to answer the question which part of phenotypic differences between parents could be expected in descendants, or among relatives (additive, dominant or epistatic), we need to count heritability in narrower sense. Geneticists who are occupied by quantitative genetic analysis have developed different methods for counting heritability coefficients, and each of them requires a specific experimental design. One of them is the regression method.

     Previous investigations the show that body size is positively correlated with mating success (Taylor, 1987;  Partridge et al., 1987;  Stamenković-Radak, 1992;  Stanić and Marinković, 1998). It has been  established that acoustic stimuli play a very important role in reproductive behavior of Drosophila (Spieth, 1952;  Von Schilcher, 1976).  Bastoc (1956), performing comparison of sex behavior of male of yellow mutant and wild type, has found that reproductive stimulation of females to a remarkable level attains the count of vibrations of wings of  males. In  courtship time, yellow males vibrate by wings with lower intensity and through greater intervals than males of wild type. That leads to significant decrease of reproductive ability of mutants. Ewing (1977) reveals analogue facts on a line characterizing the level of reduction of the wing plate.

     In numerous papers, morphometric analysis of Drosophila melanogaster individuals has been performed to understand sexual and reproductive selection better. This form of selection occurs when one phenotypic aspect of the organism varies, influencing mating success and inducing competition within the same sex (Wilkinson, 1987). In natural habitats the wing size was larger in mated D. melanogaster males in comparison with those that did not mate with females (Taylor and Kekić, 1988). Also, body size dependent assortative mating was demonstrated in laboratory lines. This success in mating of males with longer wings could be explained by stronger acoustic stimulation, greater competition ability, and the fact that larger males live longer.  Taking into consideration all of this, we decided to perform quantitative analysis of wing length in Drosophila melanogaster using regression method for a determination of the heritability coefficient of this character.

     The experiment was performed with individuals of Canton-S wild type and sepia (se) individuals of a mutant line of Drosophila melanogaster purchased from the Umea Stock Center. The  flies have been maintained on standard nutritious medium for Drosophilidae (cornmeal, sugar, agar, yeast) in termostat, under optimal conditions, at 25^oC and humidity of 60%. Since development has been conducted under optimal conditions, new individuals of wild type were hatched from eggs 12-15 day after laying eggs, while mutant line individuals were not hatched until the 20^th day.

     Initially, we have formed groups of close relatives (parents - descendants) in both lines, and after that proceeded with morphometric analysis of all individuals (P and F₁ generation) according to the method of Partridge (1987). The wing of each individual was taken and placed on a microscopic glass slide in a drop of 70% ethanol, and using the ocular with a measuring scale, the distance from the cross of the frontal transversal and longitudinal vein, to the spot where longitudinal vein joins the distal edge of the wing was determined. The wing length was expressed in measurement units on an ocular scale, where 38 units = 1mm. On the basis of covariances of individuals, the values of h_n² were determined separately for individuals of the male and female sex in both lines (h_n² = S_a²/S_f²). On the basis of comparisons of obtained results our conclusions are presented in the text to follow.

      By calculating the regression coefficient of one or both parents on descendants of male or female sex, similarity between close relatives was determined; in that way heritability in narrower sense for wing length in individuals of wild type, as well as in the mutant sepia line, was determined. The method of controlled crossing to obtain the data on which h² was determined has allowed us to have information about maternal effect in Drosophila.

     The results we obtained show (Tables 1a and 1b) that the additive genetic component contributes to a greater similarity between parents and female progeny in mutant sepia line, while in individuals of wild type that component of genetic variability is more expressed between parents and male progeny. On the basis of this, we assume that rather nonadditive effects could explain phenotypic variance among individuals of wild type Drosophila, i.e. by mechanisms marked as epistasis and dominance (dom-rec. interactions).  As can be seen, the mutation process clearly influences the components of phenotypic variability of quantitative traits, which is in agreement with the original hypothesis of Sewal Wright  (Wright, 1932).

     In regard to the importance of genetic variance components, the results indicate the following:

1.    In wild type Drosophila - additive between parents and descendants of male sex; dominance and epistasis between parents and descendants of female sex;

2.             In individuals of the se line - additive between parents and descendants of female sex; dominance and epistasis between parents and descendants of male sex with maternal effect.

     Regression coefficients, and in agreement with them the heritability coefficient in narrower sense, are completely different in individuals of the mutant line, in relation to wild type of Drosophila. Variability of wing length in individuals of male sex of wild type Drosophila is conditioned by additive action of genes (f - 0.13; m - 0.29).  Opposite to these data, in individuals of the analyzed mutant line, part of the additive action of genes is remarkably higher in female sex (0.23; 0.50) than in male sex (0.07; 0.08). For  variability  of wing length of se line males, interactions between allelic and nonallelic genes are more important than cumulative action of those genes.

     These differences in heredity of wing length we may be explained by cytoplasmic-genetic form of maternal effect. The highest heredity was established in pairs of mothers-female progeny of the se line; that could be explained by influence of additive action of nuclear genes, additive action of cytoplasmic (mitochondrial) genes, and interactions between nuclear and cytoplasmic genes.

     It means in the analyzed mutant line that the maternal effect is especially expressed between ²mothers² and female progeny, while in the wild type of Drosophila the influence of maternal  extra chromosomal effects is more expressed in female progeny. Based on that fact we generally conclude that:

-       mutations of genes determining qualitative traits influence also variability of quite different quantitative traits;

-       mutations appear as a factor that causes a conversion of epistatic interaction in additive genetic variance, that is in accordance with the assertion of some former findings (Wade and Goodnight, 1998).

     This mechanism exists in relatively small populations, and it accelerates the process of genetic differentiation, which is one of the first steps in the process of speciation. Our results suggest the presentation of total genetic variance in the form of the next equation: V_G = V_A + V_D + V_C + V_AC + V_DC, were the V_A and V_D are additive and dominant components of genetic variance of genes from nucleus, V_C is variance of additive effects of cytoplasmatic genes, and V_AC and V_DC are variances of interactions between genes of nucleus and cytoplasm (Lynch and Walsh, 1998). Also, the regression coefficient between fathers and their progeny is modified because of genetic covariance between direct and maternal effects, since the similarity between ''father'' and his descendants is influenced by genes from the nucleus, as well as by genes that have origin from his mother and such maternal effects. Cited facts should be taken into account when differences in heritability between laboratory and field estimations are considered (Weigensberg and Roff, 1996), as well as the influence of additive genetic variance on adaptive evolution of natural populations.

     KEY WORDS: D. melanogaster, sepia, heritability, wing length, regression, maternal effects.

     References: Ewing A.W., 1977, in How Animals Communicate (T.A. Seboek);  Lynch M., and B. Walsh 1998, Sinauer Assoc., Sunderland;  Partridge L., et al., 1987, Anim. Behav. 35;  von Schilcher, F., 1976, Anim. Behav. 24;  Spieth, H.T., 1952, Bulletin of the American Museum of Natural History 99;  Stamenković-Radak, M., 1992, Ph. D. thesis, Belgrade;  Stanić S., and D. Marinković 1998, Genetika 30, Belgrade;  Stanić, S., and D. Marinković 1999, Arch. Biol. Sci. 51, Belgrade;  Taylor, C.E., 1987, The American Naturalist 129;  Taylor, C.E. and V. Kekić 1988, Evolution 42;  Wilkinson, G.S., 1987, in Sexual Selection – Testing the Alternatives; Wright, S., 1932, Proc.6^th Intern. Congr. Genet.1;  Wade, M.J., and C.J. Goodnight 1998, Evolution 52;  Weigensberg, I., and D.A. Roff 1996, Evolution 50.